The primary goal of this R01 proposal is to study the impact of varying supervision strategies on clinician fidelity and client outcomes in a community-based setting. Prior research has established that training approaches that do not include a period of intervention-specific supervision or consultation are ineffective and that implementation efforts that include only an initial period of supervision show an eventual attenuation of gains in knowledge and fidelity in practice. Ongoing supervision may be required for effective and sustained implementation of evidence-based practices (EBPs) in community-based settings. However, supervision is one of the least investigated aspects of training. Gold standard elements of supervision from efficacy trials include review of sessions, standardized procedures for monitoring client outcomes and model fidelity, and ongoing skill-building (e.g., behavioral rehearsal). The degree (e.g., frequency, intensity) to which these strategies are used in community-based settings is unknown. There are a growing number of national and statewide efforts to increase the reach of EBPs through dissemination and implementation initiatives. There are 18 statewide initiatives to implement Trauma-focused Cognitive Behavioral Therapy (TF-CBT), an EBP for child trauma exposure and sequelae. The five-year, state- funded, Washington State TF-CBT Initiative is representative of many of these efforts. Many of the community- based TF-CBT implementation efforts, and those for other EBPs, include a specific focus on supervisors. However, the limited scientific literature provides very little guidance for these efforts. Aims of the current trial include 1) studying supervision with existing implementation supports; particularly presence of gold standard elements; 2) evaluating the effects of varying supervision strategies on fidelity and client outcomes; and 3) testing the mediating effect of treatment fidelity on the relationship between supervision type and client outcomes. We propose a two-phased, within-subjects and between subjects design. In Phase I (9 months), we will examine supervision with implementation support. In Phase II (30 months), we will examine two specific supervision conditions, each including varying EBP supervision elements. We plan to enroll 20 supervisors and 140 clinicians, with randomization to condition occurring at the clinician level (i.e., all supervisors will provide all conditions). We will enroll 1 youth per clinician in Phase I (N = 140) and 3 cases per clinician in Phase II (N = 420). The findings will yield important recommendations for supervision strategies that are effective in promoting high-fidelity implementation and are most feasible in community-based settings.